The goal of the proposed research is to expand the capabilities and performance of stochastic single- molecule superresolution (SSMS) microscopy through the development of improved fluorescent protein variants. A primary objective of this project will be to enable SSMS imaging in living cells of three or more proteins simultaneously, allowing visualization of multi-protein complex dynamics in a biologically relevant context with single-molecule localization precision of 20 nm or less. Currently, live-cell SSMS imaging is largely limited to one- and two-color labeling, due to the lack of well-performing fluorescent protein variants at appropriate wavelengths. Much of this shortfall in suitable fluorescent protein labels has come from a reliance on fluorescent proteins from a very limited number of species as starting materials for engineering improved variants. This project will take advantage of wild-type fluorescent proteins with novel optical properties to be cloned from a broad range of species, including several Great Barrier Reef corals made available through an international collaboration. The use of this wide array of novel starting materials will maximize the odds of successful engineering of monomeric fluorescent proteins with markedly improved performance over current technology. Specific emphasis will be placed on functional screening in mammalian cells, rather than directed evolution exclusively in bacteria, as has been the standard approach to most past fluorescent protein engineering efforts. This expanded use of image-based screening in mammalian cells will facilitate the selection of variants that perform optimally in th primary end-user applications of this technology. The end products of this project will be complete sets of monomeric fluorescent protein variants in three classes: (1) standard (non-photoactive) fluorescent proteins in the orange, red, and far-red wavelength classes that display improved brightness, photostability, performance as fusion tags, and performance as SSMS imaging tags; (2) photoactivatable and photochromic fluorescent proteins across the entire visual spectrum that display equivalent performance to standard fluorescent proteins, but also include exceptionally high contrast between dark and activated states; (3) photoconvertible proteins, focused on the development of novel wavelength classes including red-to-yellow, blue-to-red, and red-to-far-red, with performance goals similar to the other two classes. For each of these classes, the goal will be development of high-performing variants that allow imaging of at least three distinct protein species simultaneously in living cells, as measured by several biologically relevant assays using SSMS microscopy. The fluorescent protein variants created in the course of this project will also be highly useful as general fluorescent tags, and are expected to be among the brightest and highest- performing fluorescent proteins yet developed.